The RECO AI rules Ontario real estate agents now need to follow come down to three core obligations: disclose when AI materially shapes a listing or advertisement, never publish an AI-altered image that misrepresents a property's physical condition, and ensure every client-facing communication still meets TRESA's honesty and accuracy standards. If you do those three things, you are largely on the right side of the guidance. The sections below explain what each obligation means in practice.
What exactly do the RECO AI rules say about listing descriptions?
RECO's guidance treats AI-generated listing copy the same way it treats any other advertising: the registrant is fully responsible for accuracy, and no tool or vendor absorbs that responsibility. If an AI drafts a description that overstates a basement's ceiling height, calls a wood-burning fireplace "gas," or omits a known material latent defect, the agent — not the software — is in breach. The practical implication is that AI copy must be reviewed and verified against the property data before it goes live, every single time.
This is not a new principle — TRESA has always held registrants to an accuracy standard — but RECO's guidance applies it explicitly to AI output, which many agents had been treating as a first draft they could publish with minimal review. That approach is now clearly on the wrong side of the line. Build a workflow that treats AI copy as a starting point, not a finished product. Our guide to writing listing descriptions that sell walks through a practical editing checklist you can bolt onto any AI tool.
How do the RECO AI rules apply to listing photos and virtual staging?
RECO's guidance requires that AI-altered or AI-generated images must not create a false impression of a property's physical condition. Specifically, virtually staged rooms must be labelled so buyers understand the furniture is not present, digitally removed items — clutter, vehicles in a driveway — that a buyer would reasonably expect to see must not be hidden, and any structural change rendered by AI (a cracked wall removed, a water-stained ceiling corrected) is considered misrepresentation regardless of how minor the edit appears.
The labelling requirement is the piece agents most commonly miss. A caption that reads "virtually staged" or "digitally enhanced" on each altered image satisfies the guidance; a blanket disclaimer buried in the listing remarks likely does not. MLS rules may impose their own labelling standards on top of RECO's, so confirm the exact wording with your board before the listing goes live. If you use AI photo tools, the Qlarify Photo Studio includes labelling prompts built into the export workflow so the disclosure travels with the image file, not just the listing sheet.
For a broader look at where AI photo editing sits within Canadian advertising rules, our post on AI listing photo rules every Canadian agent must know covers the national picture alongside the Ontario-specific guidance.
Does RECO's guidance change how agents must handle AI in client advertising and social media?
Yes. Under RECO's guidance, AI-generated advertising — including social posts, email campaigns, and paid ads — is held to the same TRESA standard as manually written material. That means no unsubstantiated superlatives ("best-priced home in the neighbourhood"), no testimonials the agent cannot verify, and no implied guarantees of outcome. If an AI content tool generates a caption claiming a neighbourhood has the "fastest-appreciating values in the city," the agent publishing it needs to be able to back that claim with current, sourced data — or the claim needs to come out.
The guidance also touches on AI-powered ad targeting. RECO's position aligns with existing human rights obligations: targeting or excluding audiences based on protected grounds — including family status, national or ethnic origin, religion, sex, or disability — is prohibited regardless of whether a human or an algorithm made the selection. If your ad platform uses automated audience optimization, audit its settings to confirm it is not inadvertently sorting on a protected characteristic. This is an area where the rules are still developing, so we strongly recommend confirming current specifics with RECO directly and reviewing the guidance with your brokerage's designated manager before running automated campaigns.
For agents building social content with AI assistance, the just-listed and just-sold social media playbook shows how to structure posts that are both engaging and defensible under TRESA's accuracy requirements.
What records do agents need to keep when AI is involved in a transaction?
RECO's guidance does not create a separate AI-specific record-keeping category, but it reinforces that agents must be able to demonstrate the basis for any material claim made in a listing or advertisement. In practice, that means retaining the source data you used to verify AI-generated copy — the Feature Sheet, disclosure documents, the MLS data input sheet — in the same transaction file where you would keep any other supporting document. If a complaint arises and the description contains an error, "the AI wrote it" is not a defence RECO will accept.
Good documentation habits matter beyond AI compliance. The FINTRAC compliance checklist for Canadian agents is a useful companion here: the same discipline that keeps your identity-verification records clean will serve you equally well for AI-related audit trails.
How should agents adjust their day-to-day workflow?
The guidance does not prohibit AI tools — it sets the conditions under which those tools can be used responsibly. A straightforward three-step review process covers most of the compliance surface area. First, verify every factual claim in AI-generated copy against the original property documents before publishing. Second, label every AI-altered or virtually staged image, individually, in the image caption or file name, and confirm your board's specific labelling requirement. Third, review AI-generated advertising copy for unsubstantiated claims before the ad goes live, particularly anything involving market performance, property values, or neighbourhood comparisons.
If you are integrating AI tools more broadly into your business — for client communication, CMA preparation, or lead follow-up — our overview of what to use AI for in real estate maps the low-risk, high-value applications against the ones that need more careful handling. The Qlarify SAM Chat assistant is built with these compliance considerations in mind, generating client-ready drafts that flag unverified claims rather than publishing them silently.
One important note on scope: RECO's guidance applies to Ontario registrants, and other provincial regulators are at different stages of developing their own AI frameworks. If you are licensed in more than one province or territory, check with each regulator separately. Even within Ontario, the guidance is relatively new and may be updated as AI capabilities evolve. Treat everything here as a working summary, not legal advice, and review the guidance document itself — along with your brokerage's AI policy — before making changes to your practice. When in doubt, call RECO's registrant line directly; they are staffed to answer exactly these questions.
TRESA's broader structural changes are the regulatory backdrop for all of this. If you have not yet worked through what Phase 3 means for your obligations as a registrant, our TRESA Phase 3 breakdown covers the provisions most likely to affect day-to-day practice alongside the AI guidance.